AI/ML Workload Solutions on Lenovo ThinkSystem V4 Servers

The evolution of artificial intelligence / machine learning (AI/ML) provides different value streams across enterprise business and drives significant business impact. Growing adoption of generative AI in the enterprise is driving the need for more hardware and accelerators which increase Total Cost of Ownership. The Lenovo ThinkSystem V4 server portfolio powered by Intel® Xeon® 6 processors with E-cores enables right-sized AI compute with efficient, secure, workload optimized solutions for all classical machine learning and enterprise private AI use cases.

In September 2024, Lenovo will start shipping the ThinkSystem SD520 V4 1U half width servers followed by the ThinkSystem SR630 V4 1U rack servers in November 2024 equipped with 6th generation Intel Xeon 6 processor family with E-Cores (codename Sierra Forest). These E-core processors have a scalable architecture with a higher number of cores to support AI/ML and enterprise workloads. The SD520 V4 servers are built with Neptune Liquid cooling modules to provide efficient cooling on the CPUs.

Figure 1. Lenovo ThinkSystem SD520 V4

Figure 2. Lenovo ThinkSystem SR630 V4

Intel Xeon 6 processors with E-cores are enhanced to deliver density-optimized compute in the most power-efficient manner. Xeon processors with E-cores provide best-in-class power-performance density, offering distinct advantages for cloud-native and hyperscale workloads:

• 2.5x better rack density and 2.4x higher performance per watt.
• Support for 1S and 2S servers, with up to 144c per CPU and TDP as low as 200W.
• Modern instruction set with robust security, virtualization and AVX with AI extensions.

Intel Xeon 6 processors with E cores are designed to be more power efficient and consume 30-40% less power than 5th Gen processors when servers are utilized at 40-60%. It dramatically reduces power and cooling costs and 6th Gen processors provide up to 20% more performance than previous generation processors which increases consolidation ratio for any workloads. The compute intensive AI/ML workloads benefit greatly from Xeon 6 architecture with Intel Optimized AI software libraries, and the solution reduces rack, power, and cooling cost to achieve better Return on Investment.

Lenovo ThinkSystem SR630 V4 and SD520 V4 servers have the following features to optimize AI/ML use cases:

• Sub NUMA Clustering (SNC) feature can provide improved performance for Resnet50.
• E-core 64-144c provide more energy efficiency and ideal for inference workloads and SMBs.
• Optimization support for AVX2-128 (VNNI/Int8 & Bfloat16), Accelerator ISA(AiA), 5G ISA.
• Fast upconvert for FP16 and BF16.
• Memory support for DDR5-6400 MT/s.
• GPU support - SR630 V4 (Up to 3 single width 75W GPUs) and SD520 V4 (1 single width 75W GPU).

Intel provides a comprehensive portfolio of AI development software including data preparation, model development, training, inference, deployment, and scaling. Using optimized AI software and developer tools can significantly improve AI workload performance, developer productivity, and reduce compute resource usage costs. Intel® oneAPI libraries enable the AI ecosystem with optimized software, libraries, and frameworks. Software optimizations include leveraging accelerators, parallelizing operations, and maximizing core usage.

Intel AI software and optimization libraries provide scalable performance using Intel CPU and GPU. Many of the libraries and framework extensions are designed to leverage CPU to provide optimal performance for machine learning and inference workloads. Intel Xeon 6th Gen Scalable processors with E-cores are compatible with many Intel Optimized AI Libraries and tools and provide ecosystem for model development and deployment for enterprise-wide use cases.

Lenovo ThinkSystem V4 systems with Intel Xeon 6 processors with E-cores and low-end GPU accelerators provide a cost effective infrastructure solution to scale your AI deployment and Generative AI use cases. With higher core counts, power efficiency and Optimized AI software, many AI/ML classical use cases and inference workloads can seamlessly run on the CPU without need for expensive GPU accelerators. 

Lenovo ThinkSystem V4 servers are supported by the XClarity One platform, a hybrid cloud-based unified systems management solution. XClarity One provides three predictive failure analytics engines to swiftly identify potential issues and minimize system downtime while increasing accuracy.

Lenovo is a US$70 billion revenue Fortune Global 500 company serving customers in 180 markets around the world. Focused on a bold vision to deliver smarter technology for all, we are developing world-changing technologies that power (through devices and infrastructure) and empower (through solutions, services and software) millions of customers every day.

For More Information

To learn more about this Lenovo solution contact your Lenovo Business Partner or visit: https://www.lenovo.com/us/en/servers-storage/solutions/ai/

References:

Lenovo ThinkSystem SD520 V4: https://lenovopress.lenovo.com/ds0184

Lenovo ThinkSystem SR630 V4: https://lenovopress.lenovo.com/ds0185.pdf

Intel AI Development Software: https://www.intel.com/content/www/us/en/newsroom/news/intel-unveils-future-generation-xeon.html

Intel Unveils Future Generation Xeon Architecture: https://learn.microsoft.com/en-us/sql/sql-server/what-s-new-in-sql-server-2022?view=sql-server-ver16